You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Copy file name to clipboardExpand all lines: articles/virtual-machine-scale-sets/virtual-machine-scale-sets-instance-flexibility.md
+7-7Lines changed: 7 additions & 7 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -16,7 +16,7 @@ ms.reviewer: jushiman
16
16
Instance Flexibility enables you to specify multiple different Virtual Machine (VM) sizes in your Virtual Machine Scale Set with Flexible Orchestration Mode, and an allocation strategy to further optimize your deployments.
17
17
18
18
Instance Flexibility is best suited for workloads that are flexible in compute requirements and can be run on various different sized VMs. Using Instance Flexibility you can:
19
-
- Deploy a heterogeneous mix of VM sizes in a single scale set up to the instance size limit. You can view max scale set instance counts in the [documentation](./virtual-machine-scale-sets-orchestration-modes.md#what-has-changed-with-flexible-orchestration-mode).
19
+
- Deploy a heterogeneous mix of VM sizes in a single scale set. You can view max scale set instance counts in the [documentation](./virtual-machine-scale-sets-orchestration-modes.md#what-has-changed-with-flexible-orchestration-mode).
20
20
- Optimize your deployments for cost or capacity through allocation strategies.
21
21
- Continue to make use of scale set features, like [Spot Priority Mix](./spot-priority-mix.md), [Autoscale](./virtual-machine-scale-sets-autoscale-overview.md), or [Upgrade Policies](./virtual-machine-scale-sets-set-upgrade-policy.md).
22
22
- Spread a heterogeneous mix of VMs across Availability Zones and Fault Domains for high availability and reliability.
@@ -42,7 +42,7 @@ The `sku.tier` property is currently an optional scale set property and will be
42
42
The `sku.capacity` property continues to represent the overall size of the scale set in terms of the total number of VMs.
43
43
44
44
### scaleInPolicy
45
-
The optional scale-in property isn't needed for VMSS deployments using Instance Flexibility. When scaling in, the scale set utilizes the allocation strategy to inform the decision on which VMs should be scaled in. For example, when using`LowestPrice`, the scale set scales in by removing the more expensive VMs first.
45
+
The optional scale-in property isn't needed for scale set deployments using Instance Flexibility. During scaling in events, the scale set utilizes the allocation strategy to inform the decision on which VMs should be scaled in. For example, when you use`LowestPrice`, the scale set scales in by removing the more expensive VMs first.
46
46
47
47
## New scale set properties
48
48
### skuProfile
@@ -55,23 +55,23 @@ The `vmSizes` property is where you specify the specific VM sizes that you're us
55
55
Instance Flexibility introduces the ability to set allocation strategies for your scale set. The `allocationStrategy` property is where you specify which allocation strategy you'd like to use for your Instance Flexible scale set deployments. There are two options for allocation strategies, `lowestPrice` and `capacityOptimized`. Allocation strategies apply to both Spot and Standard VMs.
56
56
57
57
#### lowestPrice (default)
58
-
This allocation strategy is focused on workloads where cost and cost-optimization are most important. When evaluating what VM split to use, Azure looks at the lowest priced VMs of those specified. Azure also considers capacity as part of this allocation strategy. When using the `lowestPrice` allocation strategy, the scale set deploys as many of the lowest priced VMs as it can, depending on available capacity, before moving on to the next lowest priced VM size specified.
58
+
This allocation strategy is focused on workloads where cost and cost-optimization are most important. When evaluating what VM split to use, Azure looks at the lowest priced VMs of the VM sizes specified. Azure also considers capacity as part of this allocation strategy. When using the `lowestPrice` allocation strategy, the scale set deploys as many of the lowest priced VMs as it can, depending on available capacity, before moving on to the next lowest priced VM size specified.
59
59
60
60
#### capacityOptimized
61
61
This allocation strategy is focused on workloads where attaining capacity is the primary concern. When evaluating what VM size split to deploy in the scale set, Azure looks only at the underlying capacity available. It doesn't take price into account when determining what VMs to deploy. Using `capacityOptimized` can result in the scale set deploying the most expensive, but most readily available VMs.
62
62
63
63
## Cost
64
-
Following the scale set cost model, usage of Instance Flexibility is free. You'll continue to only pay for the underlying resources, like the VM, disk, and networking.
64
+
Following the scale set cost model, usage of Instance Flexibility is free. You continue to only pay for the underlying resources, like the VM, disk, and networking.
65
65
66
66
## Limitations
67
-
- Instance Flexibility is currently available in the following regions: West US, West US2, West US3, East US, East US2, CentralUS, South Central US, North Central US, West Europe, North Europe, UK South, and France Central. More regions will be supported during Public Preview.
67
+
- Instance Flexibility is currently available in the following regions: West US, West US2, West US3, East US, East US2, CentralUS, South Central US, North Central US, West Europe, North Europe, UK South, and France Central.
68
68
- Instance Flexibility is only available for scale sets using Flexible Orchestration Mode.
69
69
- Instance Flexibility is currently only available through ARM template and in the Azure portal.
70
70
- You must have quota for the VM sizes you're requesting with Instance Flexibility.
71
71
- You can specify **up to** five VM sizes with Instance Flexibility at this time.
72
72
- Existing scale sets can't be updated to use Instance Flexibility.
73
73
- VM sizes can't be changed once the scale set is deployed.
74
-
- For REST API deployments, you must have an existing virtual network inside of the resource group that you'll be deploying your scale set with Instance Flexibility in.
74
+
- For REST API deployments, you must have an existing virtual network inside of the resource group that you're deploying your scale set with Instance Flexibility in.
75
75
76
76
## Deploy a scale set using Instance Flexibility
77
77
The following example can be used to deploy a scale set using Instance Flexibility:
@@ -143,4 +143,4 @@ Lastly, be sure to specify the `skuProfile` with **up to five** VM sizes. This s
143
143
Yes, you can use both Spot and Standard VMs in your scale set deployments using Instance Flexibility. To do so, use [Spot Priority Mix](./spot-priority-mix.md) to define a percentage split of Spot and Standard VMs.
144
144
145
145
## My region doesn't support Instance Flexibility today, will it support Instance Flexibility in the future?
146
-
Instance Flexibility is rolling out to all Azure regions during Public Preview. The documentation will be updated as more regions are supported. Instance Flexibility is currently available in the following regions: West US, West US2, West US3, East US, East US2, CentralUS, South Central US, North Central US, West Europe, North Europe, UK South, and France Central.
146
+
Instance Flexibility is rolling out to all Azure regions during Public Preview. Instance Flexibility is currently available in the following regions: West US, West US2, West US3, East US, East US2, CentralUS, South Central US, North Central US, West Europe, North Europe, UK South, and France Central.
0 commit comments